期刊文献+

一种基于球状分布的SVM核选择方法

A Kernel Selection Approach Based on the Characteristics of Sphere Distribution
下载PDF
导出
摘要 核选择是支撑向量机(Support Vector Machine,SVM)研究中的核心问题之一。提出了一种基于数据分布特征的SVM核函数选择的方法。分析了几种常用核函数的性能,提出了判断数据呈球状分布的方法,探讨了SVM核函数及其参数选择与数据分布的相关性。数值实验说明了该方法的可行性与有效性。 The kernel selection is one of the key problems for support vector machine (SVM). A new way to select the kernel function is presented. It is based on the characteristics of data distribution. The paper analyses the existing kernel function and presents an approach to determine sphere distribution. And then, on the basis of determining sphere distribution, this paper discusses how to select the kernel function. The simulation experiments demonstrate the feasibility and the effectiveness of the presented approach.
出处 《办公自动化(综合月刊)》 2010年第3期38-40,共3页 Office Informatization
基金 国家自然科学基金(No.60673095 No.70471003) 山西省高校科技研究开发项目(No.200611001) 山西大学商务学院院基金~~
关键词 支撑向量机 核选择 球状分布 球面坐标核 Support vector machine Kernel selection Sphere distribution Sphere coordinate kernel
  • 相关文献

参考文献11

  • 1V. Vapnik. The Nature of Statistical Learning Theory [M]. Springer2Verlag, New York, Inc, 2000.
  • 2J.C. Burges. A Tutorial on Support Vector Machines for Pattern Recognition [M] . Kluwer Academic Publishers , Boston , 1999.
  • 3J. B. Gao, S. R. Gunn and C. J. Harris. A probabilistic framework for SVM regression and bar error estimation [J]. Machine Learning, 2002, 46: 71-89.
  • 4K.B. Duan, S. Keethi , A.N. Poo Evaluation of simple performance measure for tuning SVM hyperparameters [J]. Neurocom- puting, 2003, 51: 41-59.
  • 5V. Cherkassky, Y.Q. Ma. Practical selection of SVM parameters and noise estimation for SVM regression [J]. Neural Networks, 2004, 17:113-126.
  • 6M. Opper and O. Winther. Gaussian process classification and SVM mean field results and leave-one-out estimator. Advance in Large Margin Classifiers [C]. Cambridge, MA: MIT Press, 2000, 43-56.
  • 7C. Gavin, T. Nicola .Efficient leave-one-out cross-validtion of kernel fisher discriminant classifiers [J]. Pattern Recognition. 2003,11:2585-2592.
  • 8W.J. Wang, Z.B. Xu, W.Z. Lu and X.Y. Zhang. Determination of the spread parameter in the Gaussian kernel for classification and regression [J]. Neurocomputing, 2003,55 (3-4) : 643-663.
  • 9张 莉,周伟达,焦李成.一类新的支撑矢量机核[J].软件学报,2002,13(4):713-718. 被引量:13
  • 10Jabri M. Robust Principle Component Analysis [J]. IEEE, Tramactiom on Neural Net-work, 2000,1 (1): 289-298.

二级参考文献6

  • 1吴佑寿,赵明生,丁晓青.一种激励函数可调的新人工神经网络及应用[J].中国科学(E辑),1997,27(1):55-60. 被引量:26
  • 2Vapnik,V.The Nature of Statistical Learning Theory.New York: Springer-Verlag,1995.
  • 3Cortes,C.,Vapnik,V.Support vector networks.Machine Learning,1995,20:273~297.
  • 4Burges,C.J.C.Geometry and invariance in kernel based methods.In: Sch?lkopf,B.,Burges,C.J.C.,Smola,J.,eds.Advance in Kernel Methods--Support Vector Learning.Cambridge,MA: MIT Press,1999.89~116.
  • 5K?nig,H.Eigenvalue Distribution of Compact Operator.Basel: Birkh?user,1986.
  • 6Wu,You-shou,Zhao,Ming-sheng,Ding,Xiao-qing.A new kind of ANN based on active function and its application.Science in China (Series E),1997,27(1):55~60 (in Chinese).

共引文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部